Van Til and Non-Overlapping Magisteria

Van Til and Non-Overlapping Magisteria

The presuppositional theologian Cornelius Van Til maintains over and over that no fact can be truly known apart from its interpretation through the concept of the triune God. This, like other grandiose-sounding claims, might be supposed to contradict the idea of non-overlapping magisteria, whether those be faith and science, faith and politics, or faith and anything else. If indeed nothing can be truly known apart from divine revelation, there is no magisterium which faith does not overlap.

That this conflict is merely apparent, becomes plain when we ask what Van Til’s claim can possibly mean. That no non-believer truly knows anything at all? The force of his statement must rest on that qualifier, truly. In other words, knowledge of any fact is inadequate if it does not find its proper place in connection with other facts. And, according to Van Til, the revelation of the triune God is the only sufficient organizing principle for these facts.

This claim can only be understood tautologically. Van Til has not demonstrated the God of the Bible to be necessary to predication; he has only demonstrated that the claim that a transcendent God exists is equivalent to the claim that the universe is intelligible. If this is granted, the unbeliever may reconcile his cosmology with the believer’s, but he does not have to admit anything specifically Christian. In other words, for all his bluster, Van Til’s argument still has very far to go if he wants to make Trinity, rather than simply transcendent deity, the prerequisite to intelligibility.1

A consideration of the natural sciences will bring out the problem more clearly. Van Til argues that the scientist can proceed in his studies only on the tacit assumption that a transcendent God exists. But if this means no more than the fact that the universe is intelligible (and his argument proves no more than this), Van Til has not exposed contradiction in the unbeliever, but has shown that the disagreement was merely semantic all along.

The point of contention, then, will be on the immanence and personalism of God – that is, the person of Christ and the work of the Holy Spirit. What then is left of Van Til’s argument that the non-believer knows no facts truly?

We might separate his strong claim above from the weaker claim that without special revelation, the unbeliever has merely failed to grasp the true significance of his facts in light of the providential theory of history. If this is the meaning of the contention that the non-believer knows no facts “truly”, it is no knock-down argument. It does not involve the non-believer in necessary self-contradiction. Still, it is this weaker claim that establishes the epistemological gulf between the believer and the non-believer. Van Til argues that knowledge produced by non-believers is still intelligible to and can be used by believers because they tacitly (although illegitimately, based on their explicit premises) accept the intelligibility of the universe and hence the Christian God.

But this solution only works because Van Til treats his two claims as if they were one. The admission of an intelligible universe says nothing at all about the personalistic ends of providence – that is, a universe in which “all things work together for the good of those that are called according to His purpose”. What then entitles the believer to take scientific knowledge, acquired on the assumption of an impersonal universe, at face value?

Taking Christian epistemology seriously means there are no theological claims on public policy

This is the claim of non-overlapping magisteria: not that any fact can be “truly” known without a personalistic providence as its organizing principle, but that natural knowledge is modular, so to speak. That is, we may admit that knowledge produced by the non-believer is necessarily subjectively invalid so far as it lacks the proper (explicit) organizing principle, but the rational faculty gives this knowledge such a form that it can be easily “dropped in” to another mind and will find its place among other facts without much alteration to the fact itself. Its internal consistency and correspondence to the external world are relatively unaffected by its place in a theory of history. This, and nothing more, is the difference between Karl Popper’s “second” and “third worlds” – the “autonomy of the third world” is nothing more than the modularity of natural knowledge.2 It is for this reason that being Christian confers no special competence in surgery, or car repair, or bearing the sword.

Of course, the facts of natural knowledge can be communicated in a barer form than they are believed; no knowledge is actually held apart from an organizing principle. And this is presumably what Van Til means in stating his claims so forcefully. This is also what makes plausible the liberal idea that institutions may be, and should be, more nearly value-free than individuals. This is nothing more than the claim that politics and faith are two non-overlapping magisteria. Institutions operate in the realm of communication, not in the realm of thought: they do not have values like individuals do, they reflect the values of individuals. Such a reflection will no doubt preclude some conceivable ideas of the Good, but the more limited the institution’s scope, the more widely affirmed – the more “neutral” – its reflection may be.3

Van Til’s weaker claim also has some bearing on the question of the culpability of unbelief. Presuppositionalism argues that unbelievers could see the truth, whereas non-overlapping magisteria are implied from the fact that they cannot. Here we must draw the distinction of Jonathan Edwards between natural and moral inability. Van Til’s argument is that unbelievers are culpable because they have the natural ability to see the truth, and do not. But this does not establish any special priority for the Christian framework in other spheres, for the argument for non-overlapping magisteria depends on moral inability, an inability which originates in the will rather than opposing it. Whether or not unbelievers can see, as a practical matter they will not, without the intervention of the Holy Spirit. This is the reason that the modularity of natural knowledge should be taken advantage of wherever possible. One can fully believe in the culpability of unbelief while recognizing the practical intractability of the epistemological gulf that a belief in providence creates.

Can any fact be truly known apart from the revelation of the triune God? If this means that knowledge produced by unbelievers has no value, then the claim is absurd on its face. If, however, it refers to knowledge of the fact rather than the fact itself – if the claims are understood tautologically and subjectively, respectively – then it can raise no objection against the liberal state in particular, or the idea of non-overlapping magisteria in general.


  1. Van Til’s use of the Trinity to solve the problem of the one and the many is quite scripturally illegitimate, and at least in A Survey of Christian Epistemology, no attempt is made to defend this use at all (much less to defend it from scripture) besides the empty assertion that this is “the” theistic position.
  2. It seems safer to speak of the modularity of individual pieces of knowledge, rather than the autonomy of the “third world” as the entire class of such pieces, because one piece of knowledge may be more or less modular than another, depending on how many complementary concepts it requires in order to make sense.
  3. Indeed, so far as the state is concerned, non-neutrality is regrettable even when it is inevitable.


Apologetics, Epistemology, Trinity, Cornelius Van Til, Jonathan Edwards, Karl Popper


Facebook Google Plus Twitter Reddit StumbleUpon

Leave a Reply

More Content

About »

Hi, I'm C. Harwick, an economics PhD candidate in Virginia with an interest in monetary theory, institutional evolution, and folk music.

Care to know more? Read on »

Twitter »

Design By Cameron Harwick Powered By Wordpress Hosted By Nearlyfreespeech No Copyrights